The nematode C. elegans has powerful genetics, a well-described nervous system, and a complete genome sequence;thus, it is well suited to analysis of behavior and development at the molecular and cellular levels. In particular, the ability to functionally map the influence of particular genes to specific behavioral consequences makes it possible to use genetic analysis to functionally dissect the molecular mechanisms underlying poorly understood aspects of nervous system function. However, many genes with critical roles in neuronal function have effects on behavior that to a casual observer appear very subtle or difficult to describe precisely. Therefore, to fully realize the potential of C. elegans for the genetic analysis of nervous system function, it is necessary to develop sophisticated methods for the rapid and consistent quantitation of mutant phenotypes, especially those related to behavior. The goal of this proposed work is to develop computer vision tools for quantitatively characterizing the phenotypic patterns caused by mutations or pharmacological treatments in C. elegans. By making it possible to precisely characterize the behavioral phenotypes of mutants with abnormal locomotion or egglaying, these tools will be particularly useful for correlating specific neurotransmission defects with characteristic behavioral patterns. These analytical tools will also be used to generate a comprehensive database containing complex behavioral data on a large set of mutant strains. This database will make it possible to identify groups of mutants and pharmacological treatments that have similar effects on behavior or development. With the accumulation of increasing phenotypic data on known mutants, it should ultimately be possible to record from unknown mutant or drug-treated animals and make informed initial hypotheses about the functions of uncharacterized genes and the targets of uncharacterized drugs.